[x86] use shuffles for scalar insertion into high elements of a constant vector
ClosedPublic

Authored by spatel on Thu, Nov 8, 10:37 AM.

Details

Summary

As discussed in D54073, we have a potential regression from more aggressive vector narrowing here, so let's try to avoid that by changing build-vector lowering slightly.

Insert-vector-element lowering always does this since there's no "pinsr" for ymm/zmm:

// If the vector is wider than 128 bits, extract the 128-bit subvector, insert
// into that, and then insert the subvector back into the result.

...but we can sometimes do better for insert-into-constant-vector by using shuffle lowering.

Diff Detail

Repository
rL LLVM